Don't use MPG

Let's say you want to compare some motor vehicles. Your options consume 15, 25, 35, 45 and 55 MPG (UK). Higher is more fuel-efficient, but how do they really compare?

The average car in the UK drives 7,400 miles per year. So if we get our spreadsheets out and calculate the typical monthly fuel usage, we get:

Each 10 MPG improvement is all over the place. It follows a classic 1/x curve, but this is not intuitive for a typical consumer. So is there another way?

In most cases litres consumed is what we want to know, so it makes more sense to measure the number of litres consumed over a specified distance, such as 100 km (62 miles).

Enter L / 100 km, commonplace in Europe:

MPG
L / 100 km
15
18.8
25
11.3
35
8.1
45
6.3
55
5.1

This way around, a lower number is more fuel-efficient.

Comparing the same vehicles again, we can see that 18.8 vs 11.3 L / 100 km is a significant 7.5 L saving, compared to 6.3 vs 5.1 saving only 1.2 L. This is much easier to compare at a glance.

Making this worse, gallons are different in the UK (Imperial) vs USA, so US MPG for the same vehicle will be a different value.

To convert UK MPG to L/100km, use: 282 ÷ MPG

To convert USA MPG to L/100km, use: 325 ÷ MPG

For example the average UK petrol car gets 36 MPG:
282 ÷ 36 = 7.8 L / 100 km.

The same applies for EVs; miles / kWh (and "MPGe") have the same problems as MPG.
Use Wh / mile or Wh / km instead.